jython-bugs — Private bug notification list. Do not post to this list.

[Jython-bugs] [Bug #129415] the parsing operation of execfile()

[<no...@so...>]

Bug #129415, was updated on 2001-Jan-19 12:40
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Closed
Resolution: Works For Me
Bug Group: None
Priority: 5
Submitted by: mianko
Assigned to : nobody
Summary: the parsing operation of execfile()

Details: I got the following problem with jython-20rc1 on
java1.3/WinNT4sp6.

"ejbittest\\test.py" works.
"ejbittest\general.py" works.
"ejbittest\test.py" not work.

Do we have somre rule which we have to put '\' if '/' is reside between
same character?? 

.............
>>> execfile("ejbittest\test.py")
Traceback (innermost last):
  File "<console>", line 1, in ?
IOError: File not found - ejbittest     est.py (The filename, directory
name, or
 volume label syntax is incorrect)
.............

Follow-Ups:

Date: 2001-Jan-25 05:40
By: pedronis

Comment:
Python strings are parsed the same way in every context like C.
execfile is simply a built-in function.
In python there's no special forms/statements that parse their arguments
in special ways.
Note: differently from C in python "\j" == "\\j" not "j".


-------------------------------------------------------

Date: 2001-Jan-19 12:45
By: mianko

Comment:
Ops, Ok.
I'm sorry. It may be wrong bug report.
But I cannot still understand why we have to use
this parser(\t is tab) for file name.

-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129415&group_id=12867

[Jython-bugs] [Bug #130021] os.stat return wrong values

[<no...@so...>]

Bug #130021, was updated on 2001-Jan-25 02:01
Here is a current snapshot of the bug.

Project: Jython
Category: Library
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: karthy
Assigned to : nobody
Summary: os.stat return wrong values

Details: jython-2.0,  Sun JDK-1.3, Linux

The os.stat differs between cpython and jython. One can expect, that not
all values is available in jython, but the time fields is 1000 times larger
in jython that in cpython which makes the time library fail when operating
on the values.

Here is an example in jython and cpython for comarrison:

Jython:
>>> import os
>>> os.stat('jakarte-test')
(0, 0, 0, 0, 0, 0, 15761L, 980410501000L, 980410501000L, 0)

Python:
>>> import os
>>> os.stat('jakarte-test')
(33204, 1217412, 3, 1, 2115, 1000, 15761, 980410501, 980410501,
980410501)


Also, the size is a long, while Python return an int.

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=130021&group_id=12867

[Jython-bugs] [Bug #129958] no interface to change drive in "change path dialog box"

[<no...@so...>]

Bug #129958, was updated on 2001-Jan-24 11:07
Here is a current snapshot of the bug.

Project: Jython
Category: Installer
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: mianko
Assigned to : nobody
Summary: no interface to change drive in "change path dialog box" 

Details: When I try to change the path on "install in", the dialog box
doesn't show any interface to change drive.


For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129958&group_id=12867

[Jython-bugs] [Bug #129746] possibe bug in jython re package

[<no...@so...>]

Bug #129746, was updated on 2001-Jan-22 17:47
Here is a current snapshot of the bug.

Project: Jython
Category: Library
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: sjprocter
Assigned to : nobody
Summary: possibe bug in jython re package

Details: re.findall throws an exception when processing some strings.

Following is a code fragment which exhibits this.  

The only difference between the two strings in the example, okinput and
badinput,
is a space between the HTML tags in badinput.  Calling findall on okinput
works and
on badinput generates an exception.  

When run from C python both strings are processed without exception.

                 - Steven

--- cut here ---

import re

def test():
	imagepattern = '(?P<img><[ \t\n]*img[^>]*>)'
	framepattern = '(?P<frame><[ \t\n]*frame[^>]*>)'

	# embed directives that use a src= construction
	extractSrcTags = '(' + imagepattern + '|' + framepattern + ')'

	okinput = """<img src="foo bar"><frame src=baz>"""
	badinput = """<img src="foo bar"> <frame src=baz>"""
	
	print re.findall(extractSrcTags, okinput)
	print re.findall(extractSrcTags, badinput)

--- cut here ---

Here is the first bit of the exception:

Traceback (innermost last):
  File "<console>", line 1, in ?
  File "E:\jakarta-tomcat\webapps\python\WEB-INF\source\bug.py", line 14,
in tes
t
  File "e:\jython-2.0\Lib\sre.py", line 59, in findall
java.lang.StringIndexOutOfBoundsException: String index out of range: -1
        at java.lang.String.substring(Unknown Source)
        at org.python.modules.sre.SRE_STATE.getslice(SRE_STATE.java:1128)

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129746&group_id=12867

[Jython-bugs] [Bug #129415] the parsing operation of execfile()

[<no...@so...>]

Bug #129415, was updated on 2001-Jan-19 12:40
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: mianko
Assigned to : nobody
Summary: the parsing operation of execfile()

Details: I got the following problem with jython-20rc1 on
java1.3/WinNT4sp6.

"ejbittest\\test.py" works.
"ejbittest\general.py" works.
"ejbittest\test.py" not work.

Do we have somre rule which we have to put '\' if '/' is reside between
same character?? 

.............
>>> execfile("ejbittest\test.py")
Traceback (innermost last):
  File "<console>", line 1, in ?
IOError: File not found - ejbittest     est.py (The filename, directory
name, or
 volume label syntax is incorrect)
.............

Follow-Ups:

Date: 2001-Jan-19 12:45
By: mianko

Comment:
Ops, Ok.
I'm sorry. It may be wrong bug report.
But I cannot still understand why we have to use
this parser(\t is tab) for file name.

-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129415&group_id=12867

[Jython-bugs] [Bug #129415] the parsing operation of execfile()

[<no...@so...>]

Bug #129415, was updated on 2001-Jan-19 12:40
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: mianko
Assigned to : nobody
Summary: the parsing operation of execfile()

Details: I got the following problem with jython-20rc1 on
java1.3/WinNT4sp6.

"ejbittest\\test.py" works.
"ejbittest\general.py" works.
"ejbittest\test.py" not work.

Do we have somre rule which we have to put '\' if '/' is reside between
same character?? 

.............
>>> execfile("ejbittest\test.py")
Traceback (innermost last):
  File "<console>", line 1, in ?
IOError: File not found - ejbittest     est.py (The filename, directory
name, or
 volume label syntax is incorrect)
.............

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129415&group_id=12867

[Jython-bugs] [Bug #129368] __getitem__ expected, but not defined for java exc.

[<no...@so...>]

Bug #129368, was updated on 2001-Jan-19 03:18
Here is a current snapshot of the bug.

Project: Jython
Category: Library
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: bckfnn
Assigned to : nobody
Summary: __getitem__ expected, but not defined for java exc.

Details: >Traceback (innermost last):
>  File "FutureSourceSpider.py", line 86, in ?
>  File "c:\software\jython\Lib\urllib.py", line 68, in urlretrieve
>  File "c:\software\jython\Lib\urllib.py", line 198, in retrieve
>  File "c:\software\jython\Lib\urllib.py", line 169, in open
>  File "c:\software\jython\Lib\urllib.py", line 273, in open_http
>  File "c:\software\jython\Lib\httplib.py", line 430, in putrequest
>AttributeError: __getitem__
>
>The line throwing the error in inside an exception handler:
>        try:
>            self.send(str)
>        except socket.error, v:
>            # trap 'Broken pipe' if we're allowed to automatically
>reconnect
>>>>>   if v[0] != 32 or not self.auto_open: <<<<<
>                raise
>            # try one more time (the socket was closed; this will
>reopen)
>            self.send(str)
>



In CPython, the exception would have been a subclass of Exception, which
defines a __getitem__ method. Index 0 would then contain the error
number and index 1 an textual description. So the httplib will most
likely work on CPython.

OTOH, the java exception is just passed back from the jython socket
module. This exception can not be indexed as a list, therefore we get
the AttributeError: __getitem__. There is also no error number in a java
exception.

I'm not yet sure what the right long term solution is. Either httplib
could test that the exception is a sequence before doing any indexing.
Or the socket module could catch the java.net exception and reraise it
as a SocketException.


For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129368&group_id=12867

[Jython-bugs] [Bug #129364] Internal exception thrown for illegal listcomp code

[<no...@so...>]

Bug #129364, was updated on 2001-Jan-19 01:52
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 1
Submitted by: bckfnn
Assigned to : nobody
Summary: Internal exception thrown for illegal listcomp code

Details: [i for i in range(10)] = (1, 2, 3)

Jython 2.0 on java1.3.0 (JIT: null)
Type "copyright", "credits" or "license" for more information.
>>> [i for i in range(10)] = (1, 2, 3)
Traceback (innermost last):
  (no code object) at line 0
java.lang.InternalError: stack < 0: -1
        at org.python.compiler.Code.push(Code.java:155)
        at org.python.compiler.Code.pop(Code.java:456)

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129364&group_id=12867

[Jython-bugs] [Bug #129363] Calling overriden method from java ctor

[<no...@so...>]

Bug #129363, was updated on 2001-Jan-19 01:47
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 3
Submitted by: bckfnn
Assigned to : nobody
Summary: Calling overriden method from java ctor

Details: Methods implemented in python will not be called from the java
super class constructor. Se this mail for more details:

http://mail.python.org/pipermail/jpython-interest/2000-September/003871.html

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129363&group_id=12867

[Jython-bugs] [Bug #129320] can't delete file when file is left open

[<no...@so...>]

Bug #129320, was updated on 2001-Jan-18 15:11
Here is a current snapshot of the bug.

Project: Jython
Category: None
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: thehaas
Assigned to : nobody
Summary: can't delete file when file is left open

Details: This is related to a bug I submitted a month or so ago.  You guys
rightfully closed it, but I finally nailed down what was causing it.  There
should be at least be a note somewhere in the documentation about it.  I'm
not sure it can be fixed.

I know, technically, you should always close your files.  But what if you
don't?  In CPython, if you don't close your files when opening them in a
function, the garbage collector closes it for you.  But there has been some
inconsisencies in how Jython behaves.

Here is my test:


import os,tempfile

def writeFile(fileName,str): 
	file = open(fileName,"w")

	file.write(str)
	# I'm not closing the file on purpose

if __name__ == '__main__':

	fileName = "bugtest.tmp"
	str = "Frank Burns eats worms\n"
	writeFile(fileName,str)

	os.remove(fileName)

This works fine on CPython on Windows 2000 and Solaris as well as Jython
2.0 on Solaris, but not with Jython 2.0 on Windows 2000.  On Windows 2000,
it gives this exception:

Traceback (innermost last):
  File "bug.py", line 16, in ?
  File "C:\jython-2.0\Lib\javaos.py", line 44, in remove
OSError: [Errno 0] couldn't delete file: 'bugtest.tmp'

In reality, I think the behavior on Windows 2000 is the correct one, but
the choice is up to you.  We did some research here, and this really does
go down to the C level.  Following is an example, which we compiled with
gcc on Solaris, gcc on Cygwin, and lc (the line compiler with Visual
Studio).

#include <stdio.h>

int main() {
  FILE *f;
  f=fopen("tempfile","r");
  if(f != NULL) {
     printf("File is Open\nAttempting to Delete...");
  }
  if(remove("tempfile")!=0) {
    printf("Delete Failed\n");
  } else {
    printf("Delete Successful (Solaris Sucks)\n");
  }
}

With both versions of gcc, this file was deleted.  With lc, the delete
failed.  (Obviously, the file "tempfile" needs to be created before running
this).



Follow-Ups:

Date: 2001-Jan-18 19:28
By: thehaas

Comment:
I'm obsessed - can't you tell??

Actually, here is an article that describes the Unix side of it . .. . it
explains the IEEE standard of unlinking files.

http://www.itl.nist.gov/div897/staff/barkley/titleissues/node26.html
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129320&group_id=12867

[Jython-bugs] [Bug #129320] can't delete file when file is left open

[<no...@so...>]

Bug #129320, was updated on 2001-Jan-18 15:11
Here is a current snapshot of the bug.

Project: Jython
Category: None
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: thehaas
Assigned to : nobody
Summary: can't delete file when file is left open

Details: This is related to a bug I submitted a month or so ago.  You guys
rightfully closed it, but I finally nailed down what was causing it.  There
should be at least be a note somewhere in the documentation about it.  I'm
not sure it can be fixed.

I know, technically, you should always close your files.  But what if you
don't?  In CPython, if you don't close your files when opening them in a
function, the garbage collector closes it for you.  But there has been some
inconsisencies in how Jython behaves.

Here is my test:


import os,tempfile

def writeFile(fileName,str): 
	file = open(fileName,"w")

	file.write(str)
	# I'm not closing the file on purpose

if __name__ == '__main__':

	fileName = "bugtest.tmp"
	str = "Frank Burns eats worms\n"
	writeFile(fileName,str)

	os.remove(fileName)

This works fine on CPython on Windows 2000 and Solaris as well as Jython
2.0 on Solaris, but not with Jython 2.0 on Windows 2000.  On Windows 2000,
it gives this exception:

Traceback (innermost last):
  File "bug.py", line 16, in ?
  File "C:\jython-2.0\Lib\javaos.py", line 44, in remove
OSError: [Errno 0] couldn't delete file: 'bugtest.tmp'

In reality, I think the behavior on Windows 2000 is the correct one, but
the choice is up to you.  We did some research here, and this really does
go down to the C level.  Following is an example, which we compiled with
gcc on Solaris, gcc on Cygwin, and lc (the line compiler with Visual
Studio).

#include <stdio.h>

int main() {
  FILE *f;
  f=fopen("tempfile","r");
  if(f != NULL) {
     printf("File is Open\nAttempting to Delete...");
  }
  if(remove("tempfile")!=0) {
    printf("Delete Failed\n");
  } else {
    printf("Delete Successful (Solaris Sucks)\n");
  }
}

With both versions of gcc, this file was deleted.  With lc, the delete
failed.  (Obviously, the file "tempfile" needs to be created before running
this).



For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129320&group_id=12867

[Jython-bugs] [Bug #129252] try/except/else statements improperly compiled

[<no...@so...>]

Bug #129252, was updated on 2001-Jan-18 07:27
Here is a current snapshot of the bug.

Project: Jython
Category: Jythonc compiler
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: bme
Assigned to : nobody
Summary: try/except/else statements improperly compiled

Details: #!/usr/bin/python

if __name__ == "__main__":
        try:
                raise "me"
        except:
                print "In except..."
        else:
                print "In else..."

The above should execute the except block only, and does when run by
jython:

[core 737] /opt/jython-2.0/jython test.py
In except...

However, when byte-compiled using jythonc and run by Java, both the except
and the else block are executed:

[core 742] java -classpath ./test.jar:/opt/jython-2.0/jython.jar test
In except...
In else...



For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=129252&group_id=12867

[Jython-bugs] [Bug #128540] Importing java classes, constructing java classes

[<no...@so...>]

Bug #128540, was updated on 2001-Jan-12 03:23
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Closed
Resolution: Later
Bug Group: None
Priority: 5
Submitted by: nobody
Assigned to : nobody
Summary: Importing java classes, constructing java classes

Details: In my app I don't use CLASSPATH much 'cause it gonna be somewhat
big (about 50 jars)
Instead of this I use wildcard as argument to my app-launcher, which parses
wildcard and finds every jar i need (builds net of ClassLoaders etc.) and
Class.forName("anyClass"); works just fine in any piece of app.

Jython package manager(SysPackageManager.java) uses "java.class.path"
property alot when huntin' for jars.
In my app-launcher expand-java.class.path code was added
someth' like that 

String currPath =  System.getProperty("java.class.path");
String separator = ";";
currPath += separator + "myPackage.jar";
System.setProperty("java.class.path", currPath);

After that modification alot of 
*sys-package-mgr*: processing new(modified) jar, 'X:\a\b\c.jar'
was shown at the Jython initialization /Py.initPython();/ assuming
everithing Ok.

BUT!
"from a.b.c.d import myClass" - ImportError: cannot import name myClass
"from a.b.c.d import *" - Ok
"dir()" - names of classes with myClass among... Hm?????
"d = new myClass()" -
java.lang.NullPointerException
        at java.lang.Class.isAssignableFrom(Native Method)
        at org.python.core.PyJavaClass.init__class__(PyJavaClass.java:143)
        at org.python.core.PyJavaClass.init(PyJavaClass.java:214)
        at org.python.core.PyJavaClass.initLazy(PyJavaClass.java:78)
        at org.python.core.PyJavaClass.initialize(PyJavaClass.java:94)
        at
org.python.core.PyJavaClass.initConstructors(PyJavaClass.java:629)
        at org.python.core.PyJavaClass.__call__(PyJavaClass.java:781)
        at org.python.core.PyObject.__call__(PyObject.java:260)
        at org.python.pycode._pyx4.f$0(<console>)
        at org.python.pycode._pyx4.call_function(<console>)
        at org.python.core.PyTableCode.call(PyTableCode.java:155)
        at org.python.core.Py.runCode(Py.java:1050)
        at org.python.core.Py.exec(Py.java:1071)
        at
org.python.util.PythonInterpreter.exec(PythonInterpreter.java:124)

When I put jar containing myClass.class in CLASSPATH explicitly -
everithing perfect.

Try to put something like 

String currPath =  System.getProperty("java.class.path");
String separator = ";";
currPath += separator + "myPackage.jar";
System.setProperty("java.class.path", currPath);

at the beginning of org.python.util.jython and don't put myPackage.jar at
CLASSPATH and HAVE FUN ;)

an...@ib...

Follow-Ups:

Date: 2001-Jan-12 04:51
By: pedronis

Comment:
Hi.

Maybe this was not clear:
clearly the technique does not work in pure java (at least under sun jvm)
either.
Consider:

* M.java
public class M {

 public static void main(String[] args) {
        
  String currPath =  System.getProperty("java.class.path");
  String separator = java.io.File.pathSeparator;
  currPath += separator + "jar1.jar";     

  System.out.println(currPath);

  System.setProperty("java.class.path", currPath); 

  try {                
   Class c = Class.forName("C");
  } catch(Throwable e) {
   System.err.println(e);
  }
                                                                           
           
 }

}


* C.java (then compiled to C.class, which is put in jar1.jar")
public class C {}


Running java M one gets:
.:jar1.jar
java.lang.ClassNotFoundException: C


Setting classpath after java init has no effect (!).

regards, Samuele Pedroni.
-------------------------------------------------------

Date: 2001-Jan-12 04:12
By: bckfnn

Comment:
The "*sys-package-mgr*: processing ..." are done only to support the dir()
and "from .. import *" on java packages. The actual loading of the class is
performed by a basic Class.forName(..).

So it is no surprise that dir() works, but that import doesn't.

However, this this more of a feature request than a bug. A task have
created to describe the feature:

 "Loading java classes from other sources" (#24502)

The bug report will be closed, but rest assured that the feature is
important and that Jython will improve in this area.


-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=128540&group_id=12867

[Jython-bugs] [Bug #128540] Importing java classes, constructing java classes

[<no...@so...>]

Bug #128540, was updated on 2001-Jan-12 03:23
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Closed
Resolution: Later
Bug Group: None
Priority: 5
Submitted by: nobody
Assigned to : nobody
Summary: Importing java classes, constructing java classes

Details: In my app I don't use CLASSPATH much 'cause it gonna be somewhat
big (about 50 jars)
Instead of this I use wildcard as argument to my app-launcher, which parses
wildcard and finds every jar i need (builds net of ClassLoaders etc.) and
Class.forName("anyClass"); works just fine in any piece of app.

Jython package manager(SysPackageManager.java) uses "java.class.path"
property alot when huntin' for jars.
In my app-launcher expand-java.class.path code was added
someth' like that 

String currPath =  System.getProperty("java.class.path");
String separator = ";";
currPath += separator + "myPackage.jar";
System.setProperty("java.class.path", currPath);

After that modification alot of 
*sys-package-mgr*: processing new(modified) jar, 'X:\a\b\c.jar'
was shown at the Jython initialization /Py.initPython();/ assuming
everithing Ok.

BUT!
"from a.b.c.d import myClass" - ImportError: cannot import name myClass
"from a.b.c.d import *" - Ok
"dir()" - names of classes with myClass among... Hm?????
"d = new myClass()" -
java.lang.NullPointerException
        at java.lang.Class.isAssignableFrom(Native Method)
        at org.python.core.PyJavaClass.init__class__(PyJavaClass.java:143)
        at org.python.core.PyJavaClass.init(PyJavaClass.java:214)
        at org.python.core.PyJavaClass.initLazy(PyJavaClass.java:78)
        at org.python.core.PyJavaClass.initialize(PyJavaClass.java:94)
        at
org.python.core.PyJavaClass.initConstructors(PyJavaClass.java:629)
        at org.python.core.PyJavaClass.__call__(PyJavaClass.java:781)
        at org.python.core.PyObject.__call__(PyObject.java:260)
        at org.python.pycode._pyx4.f$0(<console>)
        at org.python.pycode._pyx4.call_function(<console>)
        at org.python.core.PyTableCode.call(PyTableCode.java:155)
        at org.python.core.Py.runCode(Py.java:1050)
        at org.python.core.Py.exec(Py.java:1071)
        at
org.python.util.PythonInterpreter.exec(PythonInterpreter.java:124)

When I put jar containing myClass.class in CLASSPATH explicitly -
everithing perfect.

Try to put something like 

String currPath =  System.getProperty("java.class.path");
String separator = ";";
currPath += separator + "myPackage.jar";
System.setProperty("java.class.path", currPath);

at the beginning of org.python.util.jython and don't put myPackage.jar at
CLASSPATH and HAVE FUN ;)

an...@ib...

Follow-Ups:

Date: 2001-Jan-12 04:12
By: bckfnn

Comment:
The "*sys-package-mgr*: processing ..." are done only to support the dir()
and "from .. import *" on java packages. The actual loading of the class is
performed by a basic Class.forName(..).

So it is no surprise that dir() works, but that import doesn't.

However, this this more of a feature request than a bug. A task have
created to describe the feature:

 "Loading java classes from other sources" (#24502)

The bug report will be closed, but rest assured that the feature is
important and that Jython will improve in this area.


-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=128540&group_id=12867

[Jython-bugs] [Bug #127422] Importing java-classes from non-defaul ClassLoaders

[<no...@so...>]

Bug #127422, was updated on 2001-Jan-03 09:30
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Closed
Resolution: Later
Bug Group: None
Priority: 2
Submitted by: nobody
Assigned to : nobody
Summary: Importing java-classes from non-defaul ClassLoaders

Details: I have my own classloader, which loads Jython also.
When I run this 

+++++++
package a.b.c
import org.python.util.PythonInterpreter;
import org.python.core.*;

public class Test{

    public static void main(String args[]) throws Exception  {
    
        PythonInterpreter interp = new PythonInterpreter();
        interp.exec("from a.b.c import Test");
        interp.exec("s = Test.countLetters('wednesday')");
        PyObject x = interp.get("s");
        System.out.println("s=" + x);
   }

    public static String countLetters(String str){
        return "Word '"+str+"' has "+ str.length() + " letters";
    }

}
++++++++++++++++
I have 
"ImportError: no module named a"

But when I explicitly write my.jar(a.b.c.Test) to classpath - everithing is
Ok.

What about usin' Thead.getContextClassLoader() in Py.findClass() or
something?

PS can't register so my e-mail is an...@ib...

Follow-Ups:

Date: 2001-Jan-12 04:05
By: bckfnn

Comment:
This feature request is described in a new task (#24502) :

 "Loading java classes from other sources"

The bug report will be closed.
-------------------------------------------------------

Date: 2001-Jan-03 12:05
By: pedronis

Comment:
more a feature request than a bug. (sorry)

notes:
 * simply making java loading through 'import' thread-context-class-loader
related: a bad confusing hack, won't do that
 * simply setting PySystemState.classLoader at init:
  not really a workaround, by now too many jythonc-related
  side-effects, e.g. sys.path loading disabled ...

This is part of the poor-man-freezing design issue
(after 2.0):

with jython -jar ... or java -jar ...

what kind of:
 python importing
 java loading
 sys-package-manager init
to offer? ...
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=127422&group_id=12867

[Jython-bugs] [Bug #128221] jpython web page should point to jython

[<no...@so...>]

Bug #128221, was updated on 2001-Jan-09 14:09
Here is a current snapshot of the bug.

Project: Jython
Category: Documentation
Status: Closed
Resolution: Fixed
Bug Group: None
Priority: 5
Submitted by: greg_ball
Assigned to : nobody
Summary: jpython web page should point to jython

Details: There are still jpython pages on sourceforge:
jpython.sourceforge.net
and sourceforge.net/projects/jpython

Neither of these point to the Jython project.

I think correcting this could help to prevent
confusion about the status of Jython.

Keep up the good work.


Follow-Ups:

Date: 2001-Jan-12 03:52
By: bckfnn

Comment:
Guido & Barry have update the SF jpython project page as well as the
ww.jpython.org and ww.python.org pages.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=128221&group_id=12867

[Jython-bugs] [Bug #128540] Importing java classes, constructing java classes

[<no...@so...>]

Bug #128540, was updated on 2001-Jan-12 03:23
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: nobody
Assigned to : nobody
Summary: Importing java classes, constructing java classes

Details: In my app I don't use CLASSPATH much 'cause it gonna be somewhat
big (about 50 jars)
Instead of this I use wildcard as argument to my app-launcher, which parses
wildcard and finds every jar i need (builds net of ClassLoaders etc.) and
Class.forName("anyClass"); works just fine in any piece of app.

Jython package manager(SysPackageManager.java) uses "java.class.path"
property alot when huntin' for jars.
In my app-launcher expand-java.class.path code was added
someth' like that 

String currPath =  System.getProperty("java.class.path");
String separator = ";";
currPath += separator + "myPackage.jar";
System.setProperty("java.class.path", currPath);

After that modification alot of 
*sys-package-mgr*: processing new(modified) jar, 'X:\a\b\c.jar'
was shown at the Jython initialization /Py.initPython();/ assuming
everithing Ok.

BUT!
"from a.b.c.d import myClass" - ImportError: cannot import name myClass
"from a.b.c.d import *" - Ok
"dir()" - names of classes with myClass among... Hm?????
"d = new myClass()" -
java.lang.NullPointerException
        at java.lang.Class.isAssignableFrom(Native Method)
        at org.python.core.PyJavaClass.init__class__(PyJavaClass.java:143)
        at org.python.core.PyJavaClass.init(PyJavaClass.java:214)
        at org.python.core.PyJavaClass.initLazy(PyJavaClass.java:78)
        at org.python.core.PyJavaClass.initialize(PyJavaClass.java:94)
        at
org.python.core.PyJavaClass.initConstructors(PyJavaClass.java:629)
        at org.python.core.PyJavaClass.__call__(PyJavaClass.java:781)
        at org.python.core.PyObject.__call__(PyObject.java:260)
        at org.python.pycode._pyx4.f$0(<console>)
        at org.python.pycode._pyx4.call_function(<console>)
        at org.python.core.PyTableCode.call(PyTableCode.java:155)
        at org.python.core.Py.runCode(Py.java:1050)
        at org.python.core.Py.exec(Py.java:1071)
        at
org.python.util.PythonInterpreter.exec(PythonInterpreter.java:124)

When I put jar containing myClass.class in CLASSPATH explicitly -
everithing perfect.

Try to put something like 

String currPath =  System.getProperty("java.class.path");
String separator = ";";
currPath += separator + "myPackage.jar";
System.setProperty("java.class.path", currPath);

at the beginning of org.python.util.jython and don't put myPackage.jar at
CLASSPATH and HAVE FUN ;)

an...@ib...

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=128540&group_id=12867

[Jython-bugs] [Bug #127679] Can't delete files with os.remove(file)

[<no...@so...>]

Bug #127679, was updated on 2001-Jan-05 08:31
Here is a current snapshot of the bug.

Project: Jython
Category: Library
Status: Closed
Resolution: Works For Me
Bug Group: None
Priority: 5
Submitted by: thehaas
Assigned to : nobody
Summary: Can't delete files with os.remove(file)

Details: I'm not sure if this is a problem with Jython, or Java but on the
Windows platform, when I try to delete a file using the os.remove function,
I get the following error:

Traceback 
  File "C:\jython-2.0b1\Lib\javaos.py", line 43, in remove
OSError: [Errno 0] couldn't delete file: 'H:\\lit_xslt\\tmp6.html'

I didn't have this happen on a Solaris box.

Here is my Java install info:
H:\lit_xslt>java -version
java version "1.3.0_01-beta"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.3.0_01-beta)
Java HotSpot(TM) Client VM (build 1.3.0-C, mixed mode)




Follow-Ups:

Date: 2001-Jan-10 01:58
By: bckfnn

Comment:
I can't reproduce this on win2k & JDK1.3. The problem can occur (on windows
only), if someone (perhaps the script itself) have the file opened.

In any case, I strongly doubt this is a problem with jython and chose to
close the bug report.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=127679&group_id=12867

[Jython-bugs] [Bug #128221] jpython web page should point to jython

[<no...@so...>]

Bug #128221, was updated on 2001-Jan-09 14:09
Here is a current snapshot of the bug.

Project: Jython
Category: Documentation
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: greg_ball
Assigned to : nobody
Summary: jpython web page should point to jython

Details: There are still jpython pages on sourceforge:
jpython.sourceforge.net
and sourceforge.net/projects/jpython

Neither of these point to the Jython project.

I think correcting this could help to prevent
confusion about the status of Jython.

Keep up the good work.


For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=128221&group_id=12867

[Jython-bugs] [Bug #127200] Incorrect code for method returning inner class instance.

[<no...@so...>]

Bug #127200, was updated on 2000-Dec-31 17:50
Here is a current snapshot of the bug.

Project: Jython
Category: Jythonc compiler
Status: Closed
Resolution: Fixed
Bug Group: None
Priority: 5
Submitted by: andv
Assigned to : bckfnn
Summary: Incorrect code for method returning inner class instance.

Details: The compiler generates incorrect java when overiding a method that
returns and instance of an inner class.  

For example.  If jython class Log inherits javax.swing.text.PlainDocument
we get this code:

public javax.swing.text.AbstractDocument$AbstractElement 
createDefaultRoot() {
        return super.createDefaultRoot();

}


Follow-Ups:

Date: 2001-Jan-07 01:12
By: bckfnn

Comment:
Fixed in proxies.py revision: 2.8;
-------------------------------------------------------

Date: 2001-Jan-01 11:03
By: bckfnn

Comment:
A quick test shows that the AbstractDocument$AbstractElement syntax is
usable with the JDK1.2.1 and the jikes compiler, but not with the JDK1.3
compiler.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=127200&group_id=12867

[Jython-bugs] [Bug #127201] Generated code overides final methods.

[<no...@so...>]

Bug #127201, was updated on 2000-Dec-31 17:58
Here is a current snapshot of the bug.

Project: Jython
Category: Jythonc compiler
Status: Closed
Resolution: Fixed
Bug Group: None
Priority: 5
Submitted by: andv
Assigned to : bckfnn
Summary: Generated code overides final methods.

Details: The compiler generates incorrect java that overides a final
method. 

For example. If jython class Log inherits javax.swing.text.PlainDocument we
get this code: 

public javax.swing.text.Position getEndPosition() {
PyObject inst = Py.jgetattr(this, "getEndPosition");
return (javax.swing.text.Position)Py.tojava(inst._jcall(new Object[] {}),
"javax.swing.text.Position");
}

which produces this error from javac:

.\jpywork\au\com\Langdale\Anti.java:606: getEndPosition() in
au.com.Langdale.Anti.Log cannot override getEndPosition() in
javax.swing.text.AbstractDocument; overridden method is final


Follow-Ups:

Date: 2001-Jan-07 01:09
By: bckfnn

Comment:
Fixed in proxies.py revision 2.7;
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=127201&group_id=12867

[Jython-bugs] [Bug #127679] Can't delete files with os.remove(file)

[<no...@so...>]

Bug #127679, was updated on 2001-Jan-05 08:31
Here is a current snapshot of the bug.

Project: Jython
Category: Library
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: thehaas
Assigned to : nobody
Summary: Can't delete files with os.remove(file)

Details: I'm not sure if this is a problem with Jython, or Java but on the
Windows platform, when I try to delete a file using the os.remove function,
I get the following error:

Traceback 
  File "C:\jython-2.0b1\Lib\javaos.py", line 43, in remove
OSError: [Errno 0] couldn't delete file: 'H:\\lit_xslt\\tmp6.html'

I didn't have this happen on a Solaris box.

Here is my Java install info:
H:\lit_xslt>java -version
java version "1.3.0_01-beta"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.3.0_01-beta)
Java HotSpot(TM) Client VM (build 1.3.0-C, mixed mode)




For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=127679&group_id=12867

[Jython-bugs] [Bug #127422] Importing java-classes from non-defaul ClassLoaders

[<no...@so...>]

Bug #127422, was updated on 2001-Jan-03 09:30
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 2
Submitted by: nobody
Assigned to : nobody
Summary: Importing java-classes from non-defaul ClassLoaders

Details: I have my own classloader, which loads Jython also.
When I run this 

+++++++
package a.b.c
import org.python.util.PythonInterpreter;
import org.python.core.*;

public class Test{

    public static void main(String args[]) throws Exception  {
    
        PythonInterpreter interp = new PythonInterpreter();
        interp.exec("from a.b.c import Test");
        interp.exec("s = Test.countLetters('wednesday')");
        PyObject x = interp.get("s");
        System.out.println("s=" + x);
   }

    public static String countLetters(String str){
        return "Word '"+str+"' has "+ str.length() + " letters";
    }

}
++++++++++++++++
I have 
"ImportError: no module named a"

But when I explicitly write my.jar(a.b.c.Test) to classpath - everithing is
Ok.

What about usin' Thead.getContextClassLoader() in Py.findClass() or
something?

PS can't register so my e-mail is an...@ib...

Follow-Ups:

Date: 2001-Jan-03 12:05
By: pedronis

Comment:
more a feature request than a bug. (sorry)

notes:
 * simply making java loading through 'import' thread-context-class-loader
related: a bad confusing hack, won't do that
 * simply setting PySystemState.classLoader at init:
  not really a workaround, by now too many jythonc-related
  side-effects, e.g. sys.path loading disabled ...

This is part of the poor-man-freezing design issue
(after 2.0):

with jython -jar ... or java -jar ...

what kind of:
 python importing
 java loading
 sys-package-manager init
to offer? ...
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=127422&group_id=12867

[Jython-bugs] [Bug #127422] Importing java-classes from non-defaul ClassLoaders

[<no...@so...>]

Bug #127422, was updated on 2001-Jan-03 09:30
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: None
Bug Group: None
Priority: 5
Submitted by: nobody
Assigned to : nobody
Summary: Importing java-classes from non-defaul ClassLoaders

Details: I have my own classloader, which loads Jython also.
When I run this 

+++++++
package a.b.c
import org.python.util.PythonInterpreter;
import org.python.core.*;

public class Test{

    public static void main(String args[]) throws Exception  {
    
        PythonInterpreter interp = new PythonInterpreter();
        interp.exec("from a.b.c import Test");
        interp.exec("s = Test.countLetters('wednesday')");
        PyObject x = interp.get("s");
        System.out.println("s=" + x);
   }

    public static String countLetters(String str){
        return "Word '"+str+"' has "+ str.length() + " letters";
    }

}
++++++++++++++++
I have 
"ImportError: no module named a"

But when I explicitly write my.jar(a.b.c.Test) to classpath - everithing is
Ok.

What about usin' Thead.getContextClassLoader() in Py.findClass() or
something?

PS can't register so my e-mail is an...@ib...

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=127422&group_id=12867

[Jython-bugs] [Bug #127340] Problem with serializable method parameters

[<no...@so...>]

Bug #127340, was updated on 2001-Jan-02 14:42
Here is a current snapshot of the bug.

Project: Jython
Category: Core
Status: Open
Resolution: Later
Bug Group: None
Priority: 5
Submitted by: njcarriero
Assigned to : nobody
Summary: Problem with serializable method parameters

Details: An object crossing from jython to java seems to lose its identity
if passed as a serializable object . The following illustrates the
problem.

==== Output from java ====
[carriero@callisto kinase-search]$ java TellMeMore
class java.util.Hashtable
{zwei=two, uno=one}
class java.util.Hashtable
{zwei=two, uno=one}

==== Output from jython 2.0b1 ====
[carriero@callisto kinase-search]$
/home/carriero/jython/jython-2.0b1/jython tmm.py
class java.util.Hashtable
{zwei=two, uno=one}
class org.python.core.PyJavaInstance
org.python.core.PyJavaInstance@38ed7d

==== Python source ====

from java.util import *
import TellMeMore

h = Hashtable()

h.put("uno", "one")
h.put("zwei", "two")

tmm = TellMeMore()

tmm.TellMeMoreO(h)
tmm.dump()

tmm.TellMeMoreS(h)
tmm.dump()

==== Java source ====

import java.io.*;
import java.util.*;

public class TellMeMore  {
    Object	o;

    public void TellMeMoreS(Serializable o) {
	this.o = o;
    }

    public void TellMeMoreO(Object o) {
	this.o = o;
    }
    
    public void dump() {
	System.out.println(o.getClass());
	System.out.println(o.toString());
    }
    
    public static void main(String args[]) {
	Hashtable h = new Hashtable();
	h.put("uno", "one");
	h.put("zwei", "two");

	TellMeMore tmm = new TellMeMore();

	tmm.TellMeMoreO(h);
	tmm.dump();
	tmm.TellMeMoreS(h);
	tmm.dump();

    }
}


Follow-Ups:

Date: 2001-Jan-03 03:20
By: bckfnn

Comment:
This problem have been raised before and we quick forgot all about it:

http://mail.python.org/pipermail/jpython-interest/1999-July/001944.html

I don't feel that the solution Jim describe is such a big hack, but it does
require changes to the highly sensitive __tojava__() methods in the
classes:

PyClass
PyFloat
PyInstance
PyInteger
PyLong
PyString

Because such a change will need a long period if test, I'll delay the fix
until after the 2.0 release.
-------------------------------------------------------

For detailed info, follow this link:
http://sourceforge.net/bugs/?func=detailbug&bug_id=127340&group_id=12867